What can the DNA in your blood reveal about your health?


(bright music) – Today on The Future of Everything, the future of detecting DNA in your blood. Now DNA is the building block of life. It is a relatively simple long molecule or polymer made out of four
components or DNA bases which have one letter
abbreviations, the famous ATCG, which stand for their chemical names. It’s like a string of beans,
beads, beads, but it is long. A human genome is made of
about three billion DNA bases, divided into 23 chromosomes. So if you add up the
beads in each chromosome, you get about three billion. You get a genome from mom
and you get one from dad. So you have two copies of the genome, mostly the same but
obviously not identical, or six billion total. Now DNA contains the blueprints
for how your cells live, how they grow, how they
interact with other cells, and like a computer
program, it allows the cell to perform simple
computations to make decisions about when and where things happen. If this goes wrong, you can get cancer. Mutations in the DNA
cause the computations and decisions to go wrong. Other things can happen too. In the last ten years,
researchers have learned that they can detect DNA in the blood. Now we knew that the cells
in the blood had DNA, so that was not surprising,
but what was surprising is that there is sometimes DNA
from other cells in the body, often cells that have died
and just released their DNA into the bloodstream. This is sometimes called cell-free DNA because it is floating in the blood and it’s not really part of a cell. Although this may seem like
it’s junk, it offers evidence of lots of other processes
going on in the body, processes diverse as cancer,
pregnancy, stress on organs, or even death and many others. Doctor Stephen Quake is a
professor of Bioengineering, Applied Physics and Physics
and Stanford University. Steve pioneered the
detection of DNA in the blood and some its first applications. Steve, what drove your
interest in detecting DNA, and what was the first demonstration that this would actually be useful? – Well, my interest came
actually when I became a father. My wife and I where in to see the doctor, and the doctor says you guys should think about getting amniocentesis. And it was seemed like
a theoretical question and something we have time to think about. We said yeah, okay, that
sounds like the right thing if recommending it.
– And this is a super risky procedure in many ways. A needle goes into the uterus near the baby to extract fluids. – Big needle right in the mom’s belly, right next to the fetus to
try to grab a few cells, and so to do genetic testing. And we said yeah, it
sounds like a good idea, thinking we schedule
another appointment for it. Next thing we knew, the
guy was turning around with a giant needle, plunges
it right into my wife’s belly, – Whoa.
– Yeah, whoa, exactly. That was our response. And it’s the response of many people who undergo that certain invasive testing. And not surprisingly,
there’s risk associated with doing that testing. Sometimes, you lose the baby
and other health problems that might happen. – How far into the pregnancy were you? – That’s typically done, I don’t know, around 14 weeks, something like that, 15 weeks, somewhere around there. And so that sensitized me to holy cow, there’s a problem here that you’re asking a diagnostic question, and there’s a lot of
risk associated with it. And so I began to think are there ways to ask these genetic
questions and do diagnostics without adding risk? And I eventually stumbled upon this old scientific literature about this cell-free DNA
that you were mentioning, which, as it turns out,
was first discovered as a phenomenon in 1948. – That’s before Watson and Crick even articulated the
importance of DNA for genetics. – It’s before the structure,
and it’s before people knew. It’s roughly contemporary
people first realized that DNA was the molecule of inheritance. – Right. – Oswald Avery just that same
year was working that out. So it was blood chemistry
to those guys who did it. But the field stayed alive, and it was mostly people
doing cancer research. And eventually, it was figured out that when you’re pregnant, some of the DNA in your blood comes from the fetus, and that was worked out in the late 1970s. And–
– And so this is not a large amount, I’m guessing, – It’s not much,
– and I know. – just a few percent of what’s there, so it’s a very challenging
measurement problem and the decade-long search
to try to figure out how to really use that
to build a diagnostic that would allow you to
understand the genetics of the baby without having
to risk the baby’s life. And we saw that at Stanford,
and it was through the work of a really terrific
graduate student in my lab when the bioengineering department
was young, Christina Fan. And that has now been the
first real clinical application of cell-free DNA in diagnostics, and that’s how I got into
it, to answer your question. – So in that initial demonstration or in your first industrial translation,
what are the things that we can actually detect from the DNA of a fetus in the mom’s blood? – Well, when we published
the paper on this, started getting press inquiries. When is this gonna be
available in the clinic? I said, I don’t know,
decades, something like that. – That’s usually the answer.
– It takes a long time, right. It turns out people jumped
on like you wouldn’t believe. Clinical trials were launched immediately. Within three years, the first real commercial
diagnostic products had been launched, and now
it’s four million women a year, something like that, get the test, and the use of
amniocentesis has plummeted. – And so now you do this as a screening before you make the
decision about the amnio. Is that the general use of it? – That was the initial indication, and it’s very quickly moving
to replacing amnio completely. – Completely, yeah. – Yeah, that’s– – And what kind of things can we diagnose in the fetus these days? – So the major genetic disorders
you have for live births are things like Down
syndrome; that’s number one. And it’s an aneuploidy is
what it’s called technically, means the extra copy of a chromosome. And there’s a few other disorders, which are extra copies of chromosomes that are also detected with this approach. – Awesome. So that has had big-time market impact, and it’s changing people’s lives. I think it’s on the street now. People know you can get this blood test instead of the amnio,
so it didn’t stop there. Now you had this hammer, and it worked. You hit one nail. What was the next nail you
guys turned your attention to? – Well, after we published that, word got around Stanford
that I was interested in non-invasive diagnostics. And I got a call one day
from Hannah Valantine, who’s a cardiologist– – Great cardiologist. – Yep, and she says, well Steve, we got a similar problem
in heart transplants. We give people a new heart,
and after the operation, we then go biopsy that new heart and rip out pieces of the tissue to make sure it’s not
being rejected by the body. And we’re doing that
every couple of months. And so is there a blood test
that could replace that? Same sort of problem, patients
were having this painful, risky procedure, and there was a question of whether it could be replaced
by a simple blood test. And so we thought about that a bit, and– – The key opportunity here is that the DNA and the heart that belongs to the donor is not gonna match the DNA of the person who received the heart, and,
like the baby and the mom, because those are different DNAs, you have a chance of picking it up. – Yeah, the key there is
that the DNA is different. A little different with
the baby and the mom because we don’t use
differences in their DNA. But in the case of the
transplant, absolutely. The whole principle is based on there being different genomes of every cell in the heart compared to other cells in the recipient’s body. And we monitor those so-called
polymorphisms, those changes. – And so you went after this,
and you were indeed able to show that people who were
in rejection were spilling, so to speak, the heart DNA into the blood, and maybe we can avoid
some of those biopsies. – Absolutely. So we did a proof of principle study with some bank samples she had, and then we wrote a grant together and were able to do a very large study on both heart and lung transplants where pretty much every transplant patient at Stanford for those two organs was enrolled in our study
over a period of three years, and were able to validate it. It was amazing. One of my kids was in
elementary school at the time, and there was a new family who
was in the class that year. And at the end of the year,
we got a note around saying that, well, there’s a
family that’s in town because they were at the
Ronald McDonald House. One of their kids was in
the hospital and very ill, and would anyone wanna put
them up for the last couple of months because their
time had run out there. And so we invited them
– Took them in. – to our house, yeah, and
very interesting family. They were immigrants from Africa. The father had been a nurse
there, had some medical training and knew that when his son
was infant and very ill that needed serious help and
eventually got him to Stanford where the son had had a heart transplant. – Whoa.
– And we were talking around the
dinner table one night, and the dad says well, and we’re just so proud to be part of this study where people are trying to figure out if they can replace the biopsies. And we enrolled our son
in it and drew the blood. I said that’s my study.
(Russ laughs) It was amazing and felt
very good about it. – Of course, of course.
– And now that’s available. So there’s now tens of
thousands of people every year who are getting that test,
and it’s saving a lot of pain and suffering for those patients. – This is The Future of Everything. I’m Russ Altman. I’m speaking with Dr. Steve
Quake about detecting DNA, and at this moment,
detecting DNA in transplant, hoping to detect rejection. So does the test detect
rejection potentially earlier than the old-fashioned
biopsy approach would? – It does, and we’ve
proved that, absolutely. You see rejection weeks,
if not a month, earlier than the biopsy. – And then presumably, that
gives the docs more option for changing the immunosuppression. – Oh, absolutely because
yeah, as you mention, all these patients are immunosuppressed to try to prevent rejection,
and too much of that, and they’ll get an infectious disease. Too little of that, you have rejection. So they can dial up the
immunosuppressants a little bit and try to avoid the rejection event, and that’s much better for the patients. Once they hit rejection, all
sorts of bad things happen, and so the whole thing is trying to keep them properly suppressed. – And just to flesh it out a little bit, how frequently are they
getting these blood draws? Is this every six months
or every three months or– – The standard of care
for the invasive biopsies was every two months, and that’s where they initially matched it. But this is the sort of thing that can and should be done more
frequently, and I think it’s gonna change the way people
treat the patients over time. – So I know that there
are more applications, and I’m interested to know
which ones you wanna talk about, but let’s talk about
one that fascinates me, which is the detection of
infectious agents in the blood. Can you tell me how this
technology has been used in that regard and what’s
the future look like? – Yeah, so when we were doing
the large transplant study, my post doc at the time, Ian De Vlaming, was looking at all the
sequencing data very carefully and realized that not
all of the sequence reads off the sequence that were
mapping to the human genome. And he said maybe 98% of it’s mapping; there’s one or 2% that aren’t. And I said that’s great. It means we’re not having
a lot of contamination and it’s all good, and he
didn’t let it go with that, thank goodness.
(Russ laughs) And he started looking at those
things that weren’t mapping, and he realized it wasn’t contamination, and they actually were not human, and it was part of the
microbiome of these individuals. So the bacteria and the
viruses and funguses that live in our body also
release cell-free DNA, and we were measuring that as well. And he realized that we could use that to monitor things like what happens to your microbiome when your
immune system gets turned off – Right, ’cause a lot of folks
– ’cause a lot of patients are immunosuppressed, exactly.
– Right. – And then we realized ’cause some of them are getting infectious disease, we could also see infectious disease. And so that has evolved into a new kind of infectious disease diagnostic, which is hypothesis free. You don’t have to test
for a particular thing. You’re essentially testing for a thousand infections all at once, and it’s just now reached
commercial development. We’re seeing the first
peer-reviewed studies showing how to use it, and it’s a
very exciting innovation for infectious disease. – So people might find this
surprising (speaks indistinctly) so let’s just unpack this a little bit. We know that there are some bacteria that live in our gut,
and we’ve always expected to see ’em there. Many of us have assumed that my blood should be pretty much infection free. That’s not where the bacteria
and the viruses live. So I guess the first question is how much of a surprise, what
do you see in normal people who are not immunosuppressed,
and how do we interpret this? Do we know that these are diseases? Are these pathogens causing problems, or might they be part of
some ecosystem of health? – Yeah, all good questions. So a fun way to think about it is to do an order of
magnitude calculation. Could we talk about calculations here? – Yes, this is something
that physicists do, folks. – So there’s a statistic going around by the microbiome people. You’ve got 10 times more bacterial cells in your body than you do human cells. If you take that at face
value and you say well, the human genome is 1,000 times longer. You said three billion base pairs, then the typical bacterial genome, which about three million base pairs. You do the math on that,
and you say by mass, all the DNA in our body is
99% human, 1% bacterial. And so if you were to mush
this all up in a blender, purify the DNA out, that’s
what would come out. – And that matches what
your post doc found. – Yes, exactly. – So are these normal signs? Are these normal organisms,
or are these things that we have to run to the doctor and get treated for?
– The vast majority of it, the vast majority of is
our normal microbiome, bugs that live with us
commensally and happy, equilibrium, with us as humans. – I’m guessing you saw viruses or bacteria that were either entirely novel or not appreciated as living in humans? – Absolutely, we have discovered traces of novel organisms that is an area of ongoing research in the group to try to understand what they are and where they fit into the tree of life. – This is The Future of Everything. I’m Russ Altman. I’m speaking with Dr. Steve Quake, and now we’re talking about
infectious disease detection. As a doctor, I know that
we have patients come into the emergency room or into the clinic with what we call FUO,
fever of unknown origin. They look sick, they have a fever, it’s not normal to have a fever, and they look infected, but
we can’t find an infection. And so I’m guessing that
one of the key applications of this technology would be, well, what DNA are we seeing in the bloodstream, ’cause that might point us
to the infectious agent. Is that how the infectious disease community
– Absolutely, yep. – is taking this up?
– Absolutely. That’s a major application. There’s a bunch of others
that are really interesting. And to come back to the
earlier point you raised about blood infections
being a different thing, the point is that the blood
is like the septic system of the body, and it’s exploring
all the tissues and organs. And when cells are dying and
they’re releasing their DNA, it picks it up and carries it. So even if the infection
is not in the blood, you see the remnants of
the infection in the blood from that cell-free DNA.
– Yes, and so the final area that I wanted to get into, of course, is cancer. And, in fact, you mentioned
cancer in your initial comments. Where are we with the detection of cancer from cell-free DNA? – Yeah, that’s been an area of
intense interest for decades. That’s the one that was
primarily driving the field before the prenatal work and because tumors have different genomes than the normal body does. And so people would monitor
those differences in the blood and try to understand how
the disease was progressing and to try to do detection. And that’s been a little
later into the clinic than the prenatal stuff,
but it’s happening now. And it’s an area
– It is. – of intense interest. There’s a bunch of companies out there that have launched tests
or about to launch tests, and it’s gonna be very important
for helping monitor course of treatment, and that’s the
first clinical application that’s out there.
– That’s what I was gonna ask. Is this about detection
or about monitoring? And it sounds like the monitoring. – That’s the first one. It’s the easiest one ’cause
you’re in such a high risk group and it makes it an easier technical task. – And you know the cancer,
so you’ve been able to characterize what
you’re expecting to find if the cancer comes back. – Correct. But the big thing to go
after is early detection, and that would help a lot of
people and save a lot of lives. And that’s something
that is gonna be coming. Maybe it’s five years, maybe it’s sooner, but there’ll be some very valuable tools for that coming down the pike; I’m pretty confident about that. – Yeah, so let’s just think
about that for a moment because one of the things
that I know is a issue is when these new technologies arise, they often move up the time of detection. So you could get the
cancer detection earlier, you get the rejection. In general, that’s a good thing. But in cancer, it’s a little tricky because there is some, if I
understand the literature, there’s some indication
that some cancers arise, and it’s the body’s own immune system suppresses the cancer
effectively before it can grow. So have people worked out what actions you should actually take if you see a very early
indication of cancer? Is it definite that we’re
gonna hit the patient very hard with chemotherapy and
radiation and whatnot, or might we still have to figure
out what to do about that? – Yeah, that’s a really good question and important issue, and
I think we’re so early on that that’s being worked out
in the clinical community. But the initial thought is not that you would go right to
treatment with chemotherapy but that you would reflex
to other testing methods that are more expensive
and more sophisticated and are not the sort of thing you use to screen people broadly
but if you got a hint that something’s wrong, you’d use them, things like imaging
techniques and such forth. – Makes sense. This is The Future of Everything. I’m Russ Altman. More with Dr. Stephen Quake about DNA, the future of health and
biology and bioengineering, next on Sirius XM Insight 121. Welcome back to The Future of Everything. I’m Russ Altman. I’m speaking with Dr. Stephen Quake about the fabulous uses of DNA that’s floating around in our cells. Now Steve, we just went through a bunch of really killer apps, but I know that there’s yet another one, which is looking at pre-term birth. And that’s a funny one to me because it’s not immediately obvious how detecting DNA would have anything to
do with a pre-term birth. So tell us that story. – Yeah, so pre-term birth
ends up being number one cause of neonatal mortality and
complications later in life. It’s a huge problem, and there’s been, despite decades of
effort, no real progress on creating a meaningful diagnostic that tells people who’s at risk. And there’s been a lot
of effort put into– – So the goal would be very early, say this looks like a pregnancy that might have some pre-term problems. – Exactly. And more generally, when
is the baby gonna be due? Even if it’s not early, can
you predict the due date? And there’s been a lot of effort put into understanding the genetics of that, the DNA base part, that
has not really had a lot of predictive power or success. And so we turned to looking at RNA, which is carries the
message from the genome and tells you about not the inheritance but the state of the cell
and body at any given point. And it turns out same guys
who discovered cell-free DNA in 1948 also discovered cell-free RNA. – They have a good year. – They did.
(men laugh) – Same year? – Same paper! – Same paper!
(men laugh) – And so we began looking at cell-free RNA as a way to measure what’s going on in the mom’s body and with
the baby and the placenta at any given point in time
and how are things changing and can that signal to us
when the baby’s gonna be born and if the baby’s gonna be born early. And we were able, after a long effort, it took seven or eight years of work by a very large group of people, a number of collaborators
here at Stanford, including David Stevenson and Gary Shaw and Yair Blumenfeld,
bunch of the MFM docs. – MFM is maternal fetal medicine
– Fetal medicine, thank you. – It’s okay, it’s my job.
– But we managed to, we managed to figure it out. And we published a paper last year showing that there’s a handful of transcripts which indicate when the mom
is gonna give pre-term birth, about two months in advance of that. – Wow, so these are like
canaries in the coal mine. – Exactly. And we found another set of transcripts which were predict gestational age, so you can tell how old the baby is and predict when it’s gonna be born. And that turns out to be a really interesting problem as well.
– I was gonna say, I thought that good old-fashioned
subtract nine months from the date of birth gets you a pretty, and in fact, I must say,
I’m born on November 5th, and that’s important because
if you go back nine months, that gets you to February
14th, Valentine’s Day. So that’s a side story. (laughs) – Okay, I got a couple
of stories there for you. – But tell me about this.
– All let me give you a couple of stories.
– Tell me about this. – So when we were having our first kid, the one with the amnio,
right, I asked the doctor what’s the due date,
tells us the due date. I said what’s the error
of your measurement, your estimate?
(Russ laughs) And he got very offended
’cause he thought I was questioning his ability as a doctor.
– Of course, of course. – We had a very tense discussion. Finally, I manage to
communicate I was asking about the uncertainty in the estimate ’cause I wanted to know when
to adjust the travel schedule to make sure I didn’t miss it. And he couldn’t tell me the uncertainty, but he told me a number that I could use to derive the uncertainty,
and so I did that. Worked out to two sigma three sigma. I had three sigma baby.
(Russ laughs) So the baby was premature
by three and a half weeks, and it was fortunate–
– Oh, it was at the border – Yeah.
– of the plus minus. – Yeah, I was fortunately in town, and fortunately, she turned out fine. But this got me aware of the importance of not only pre-term birth
but also understanding, trying to understand when
the baby’s gonna be born and prediction of due date. – Okay, so you sold me. This is actually an impactful question. – Yes, exactly.
– So what can you guys do? – Well, it’s early still. Our first paper was a
small number of women, few dozen women, and yet
it seems very promising, and we’ve now been able to reproduce it in a different cohort that
we can predict pre-term birth and gestational age and,
from gestational age, hopefully predict when the
normal baby’s gonna be born. But it’s all now going into
much larger clinical trials to validate it. It’s very much the beginning of the story, but it’s an exciting one. – So, great. This is a new molecule for our discussion, this RNA molecule, also from the baby or the placenta or both and a combination of maternal and fetal factors – Yes.
– gives you the data you need and a big data mining
approach, not to overuse that, to actually draw inferences
that might be very impactful both for the actual due date but, more importantly, for
uh oh, we have a woman who might be having a pre-term
birth, let’s do what we can, and again, the ability
of doctors to intervene is probably much better if
they have two-month warning. – Correct, correct. – Well, so this has been an amazing ride, and I wanna turn our
attention a little bit now to a separate thing but very excited that you’re involved with. A few years ago, I think three years ago, the Chan and Zuckerberg
Foundation announced the creation of a big biomedical research institute with you and a colleague
from UCSF, Joe DeRisi, as co-presidents, and it
had a very bold mission. The mission was to, I believe, cure or manage all disease
by the end of the century, something like that; you
can correct me if I’m wrong. – Cure, treat, or
prevent all human disease by the end of the century. – Bingo. So you agreed to that charge. You’ve now been doing it for three years. Can you tell us a little
bit about how it was set up and why it was set up, and
is it really even possible to imagine that level of
progress in the next century? – Yeah. So since we’ve been talking
about becoming parents, and Mark and Priscilla began
to turn their attention to philanthropy in a pretty large way when they became parents. And they wrote an open letter
before their first daughter was born that launched the
Chan Zuckerberg Initiative and ultimately the Biohub with this idea of trying to create a better world. – It’s called the Chan Zuckerberg– – Chan Zuckerberg Biohub.
– Biohub, mm-hmm. – And so in their children’s lifetime, so broadly in this hundred-year span, they wanted to see if they
could fund scientific research that would help make the
world a healthier place for their kids and everyone
else, which is a lovely mission. And it sounds crazy, right?
– Sounds crazy. – It sounds absolutely absurd, and for awhile, I couldn’t
say it even to them with a straight face. (laughs)
– And yet a few moments ago, you said it forcefully and
convincingly, so wait to go. – In, well–
– What turned you? – Well, you think about it for awhile, and it helps to think backwards in time and think about how far
medicine has advanced in the last 100 years. And in this country, mortality
has been cut in half. And the things that kill
us now are very different than the things that
killed us 100 years ago. Primarily, it was infectious disease then. Now it’s things like heart
disease and such forth. And so we’ve eliminated
entire classes of diseases, effectively, and cut mortality in half. So you can project
forward another 100 years and say if we don’t do anything, we should get another factor of two. And with some really serious effort, maybe we can do better than that. It’s just very hard for people to think on century-long timescales. We’re thinking when’s
our next grand proposal or something like that or
when’s our next student gonna graduate, and it’s not often that we have the opportunity to think on that sort of timescale. – This is The Future of Everything. I’m Russ Altman. I’m speaking with Steve Quake, now about curing, managing, treating, and what was the other verb? – Prevent.
– Preventing all disease. Do you take a portfolio approach? It sounds like you were talking about the causes of death 100 years ago, so you have to look at
the causes of death now. And I guess you have to
pick the low-hanging fruit to say how do we make progress. So how have you decided
to deploy the assets of the Biohub for the
next five to 10 years? – Yeah, well, a two-fold approach. One approach is to pick a couple of areas that capture a large part of
the global burden of disease, and we’ve chosen two that we focus on in our internal research. One is cell biology, and a lot of diseases are a consequence of
disorders of cell biology, cancer, heart disease, pulmonary disease, a number of neurological diseases. And so better understanding how cells work will lead to new therapies and treatments. – Could be a platform of discoveries that will have multiple applications. – Right, exactly, and
that covers a large part of the global burden, as you
work out the numbers of that. The other big part is infectious disease. – So it’s still a problem. – Still a problem worldwide, absolutely. There’s a bunch of open
areas, malaria, HIV. There’s a bunch of other ones, TB, number of viral infections. So that’s our other big internal effort. And at the Biohub, our
researchers have been hired to focus on those two areas. Now the other, all the rest, (men laugh) what we’ve done is we’ve partnered with the Bay Area University, to Stanford, UCSF, and Berkeley, and we fund research of nearly 100 faculty at those universities
across everything else. We have an open competition. We’ve committed roughly $100 million to those faculty over the next five years, and we’ll do it again for
the second five years. And we’re encouraging them
to work on the riskiest, most exciting ideas, whether
they’re basic science, technological, or more disease focused, to cover the span of where we think a lot of the great innovations are gonna come over the next decades. – So that does sound compelling. So basically, a two-fold strategy with some top-down projects that you know are gonna be impactful, and
then you spread your bets by giving money to a bunch of smart people and say just do what you think is right. And the hope is that that will lead to the next set of
challenges that you guys can perhaps adopt as top-down challenges. – That’s right.
– So how is it going? – Well, as you mentioned, we just celebrated our third
birthday three weeks ago. Joe and I have been working
really hard, but it feels great. I feel like we’re at full steam, and great science is happening, and the people we’re funding
are doing great work, and the future is bright. – And I guess are the donors satisfied? Are the people who put up the funds, are they starting to see
their fruits of their vision? – Well, you’ll have to ask them that. (Russ laughs)
It’s not my place to say. But they haven’t fired us yet, and so we take that as a good sign. – Thank you for listening
to The Future of Everything. I’m Russ Altman. If you missed any of this episode, listen anytime on demand
with the Sirius XM app.

Leave a Reply

Your email address will not be published. Required fields are marked *